A Nodal Method for Phase Change Moving Boundary Problems

Research output: Contribution to journalArticle


A semi-analytic numerical scheme has been developed to solve the one-dimensional, moving boundary phase change problem with time-dependent boundary conditions. Locally analytic, approximate solutions are developed for the position of the moving boundary, and for temperature distribution. Set of discrete equations are obtained by applying these solutions over space-time nodes, and by imposing continuity of temperature and heat flux. Application of this so-called nodal integral approach to the nonlinear Stefan problem shows that the scheme is O(Δx2), and that it predicts the position of the moving boundary and the temperature distribution within the domain very accurately. For example, with as little as two nodes in the spatial domain, the location of the moving boundary for the case of an exponentially increasing surface temperature on the boundary, after one dimensionless time unit, is found with an error of less than 1%. In addition to large size nodes in space, this scheme also allows the use of very large size time steps. Comparison of numerical results with reference solutions is presented.

Original languageEnglish (US)
Pages (from-to)211-221
Number of pages11
JournalInternational Journal of Computational Fluid Dynamics
Issue number3-4
StatePublished - Jan 1 1999


  • Moving boundary problem
  • Nodal integral approach
  • Stefan problem

ASJC Scopus subject areas

  • Computational Mechanics
  • Aerospace Engineering
  • Condensed Matter Physics
  • Energy Engineering and Power Technology
  • Mechanics of Materials
  • Mechanical Engineering

Fingerprint Dive into the research topics of 'A Nodal Method for Phase Change Moving Boundary Problems'. Together they form a unique fingerprint.

  • Cite this